High-temperature blocking device opening and closing timing regulation method based on production data

Abstract

This invention relates to the field of industrial process automatic control technology, specifically a method for controlling the opening and closing sequence of high-temperature blocking equipment based on production data. The invention establishes a unit response characterization based on the historical action quantities of the controlled unit and its recovery response after reactivation, and determines the current round of action requirements by combining operational deviations. It determines the safety boundary between temperature and opening / closing response along the upstream production cycle, obtains the allowable action quantity through temperature extrapolation, and selects the time domain where the allowable action quantity covers the current round of action requirements and the load state is at a local trough as candidate phases. It verifies the supply conditions based on the pressure recovery trajectory of the common execution source, configures non-overlapping admission phases according to the action duration and pressure recovery sequence, and generates opening / closing sequence control commands. The commands are executed, terminating the operation when the temperature reaches the safety boundary or the current round of action is completed, and updating the unit response characterization and the next round of action requirements based on the actual action quantity and recovery response. This invention improves the adaptability and reliability of opening / closing sequence control.

Description

Technical Field

[0001] This invention relates to the field of industrial process automatic control technology, specifically to a method for controlling the opening and closing timing of high-temperature blocking equipment based on production data. Background Technology

[0002] Multiple high-temperature blocking devices are installed in parallel on the production line. Each device has two working positions: on and off. The control system issues commands to switch between these two positions to adjust the activation and deactivation of each controlled unit. The timing, duration, and sequence of each device's action constitute the opening and closing sequence, which directly affects process operating indicators, energy consumption, and equipment availability.

[0003] There are four main types of existing control methods: timed rotation based on a set cycle; triggering by exceeding the unit's pressure difference or flow rate limits; switching modes based on the ratio of the unit's flow rate to its rated value; and setting the frequency of actions based on multiple unit measurements using fuzzy rules. Although the decision-making basis of these four methods differs, the sampling points are all located within the controlled unit itself or downstream of it, representing post-event feedback of deviations that have already occurred. In contrast, the disturbance source is located in the upstream production process and arrives at the controlled unit after transmission, so the decision-making time is necessarily lagging behind the arrival time of the disturbance.

[0004] This leads to three main problems. First, placing any device in the blocking position increases the load on other operating units. Therefore, the cost of an action occurring in the high-load range is significantly greater than that occurring in the low-load range, and the unit's own observations cannot indicate the arrival of the low-load range in advance. Second, if conduction is restored before the blocking duration is sufficient, the effect of the operation is immediately negated. The consumed working medium and the number of actions already performed cannot be recovered and must be repeated, which further exacerbates the losses. Third, the allowable range of this duration is constrained by both the state of the equipment itself and the state of the medium, and changes in real time with the operating conditions, making it difficult to adapt using a constant setting. Therefore, how to adapt the opening and closing sequence to changes in operating conditions that have not yet arrived is a technical problem that urgently needs to be solved in this field.

[0005] To address this, a method for controlling the opening and closing timing of high-temperature blocking equipment based on production data is proposed. Summary of the Invention

[0006] The purpose of this invention is to provide a method for controlling the opening and closing sequence of high-temperature blocking equipment based on production data. The method determines the action requirements based on the historical actions and recovery responses of the controlled unit, and selects and allocates the admission phase by combining temperature safety boundaries, production cycle time, load troughs and common execution source pressures, generates opening and closing sequence control commands, and updates subsequent control based on the execution results.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A method for controlling the opening and closing timing of high-temperature blocking equipment based on production data includes: Based on the historical action volume of each controlled unit and the recovery response after reactivation, a unit response characterization is established, and the action requirements for this round are determined in combination with operational deviations. The safety boundary of the upstream production cycle is analyzed to meet the temperature control state and the opening and closing response state. Combined with temperature extrapolation, the amount of action that can be performed before reaching the safety boundary is determined as the allowable action amount. The time domain that covers the current action demand and the load state is at a local trough is determined as the candidate phase. Based on the pressure recovery trajectory of the public execution source, the supply conditions of the candidate phase for the current action demand are verified, and the candidate phase that passes the supply verification is determined as the admission phase; According to the action period and pressure recovery sequence, each controlled unit is configured with a non-overlapping access phase to form and output opening and closing timing control commands. Execute the opening and closing timing control command, terminate the current operation phase when the measured temperature reaches the safety boundary or the current round of action is completed, and update the unit response characterization and the next round of action requirements based on the actual amount of action and the recovery response after re-entry.

[0008] Preferably, the process of obtaining the unit response characterization includes: taking the time of re-entry as the boundary, within a continuous time domain in which the upstream production cycle has not been switched, there are no opening or closing actions, and the total driving pressure difference is within the historical comparable range of the production stage, continuing the evolution of the total flow characterization before re-entry to form a local reference; determining the transition of the total flow characterization after re-entry relative to the local reference as the recovery response; and associating the recovery response with the actual action quantity of the corresponding operation phase according to the control cycle to form a unit response characterization that characterizes the correspondence between the action quantity and the flow recovery.

[0009] Preferably, the process of obtaining the current round of action requirements includes: extending the evolution of the operating deviation formed by adjacent control cycles in segments along the upstream production cycle to obtain the estimated time when the operating deviation reaches the upper limit of the allowable operating range; reserving the shutdown response time and the time required for the current round of operation ahead of the estimated time, and determining the latest shutdown command time as the action deadline; taking the estimated value obtained by extending the evolution of the operating deviation to the completion time of the current round of operation as the deadline deviation state; determining the target recovery degree by the sum of the expected growth between the deadline deviation state and the earliest time when the controlled unit can obtain the next operation phase from the completion time of the current round of operation, relative to the upper limit of the allowable operating range; searching the unit response representation in order of increasing action quantity to the position where the recovery response is first not less than the target recovery degree; when there is an incomplete action quantity, taking the position corresponding to the incomplete action quantity as the search starting point, and determining the action quantity corresponding to the search position as the current round of action requirements.

[0010] Preferably, the process of obtaining the allowable action quantity includes: determining the safety boundary based on the temperature control lower limit corresponding to the medium state and the protection margin corresponding to the opening and closing response hysteresis; for each candidate time period divided by the stage switching position of the upstream production cycle, taking the body temperature evolution before the start of the candidate time period and in the same production stage as a local reference, and forming a no-action temperature baseline along the local reference; in each control cycle during the historical action period, taking the temperature difference between the no-action temperature baseline and the measured body temperature as the temperature margin consumption, and associating the temperature margin consumption with the action process, when there are multiple valid records under the same conditions, taking the larger consumption point one by one to form a temperature margin consumption trajectory; deducing sequentially from the starting temperature of the candidate time period along the temperature margin consumption trajectory, and determining the action quantity that can be executed before reaching the safety boundary as the allowable action quantity.

[0011] Preferably, the process of obtaining the candidate phase includes: taking the upstream event formed by the stage switching signal or the change in the feeding rate as the starting point, searching in the historical records for the moment when the total flow and the body temperature show a corresponding turning point, and obtaining the propagation delay of the production stage; shifting the historical temperature state evolution and load state evolution within the same production stage and whose stage start conditions are consistent with the current candidate time period according to the propagation delay to the corresponding candidate time period, and replacing the starting point of the migrated trend with the starting state of the candidate time period, and sequentially deriving the corrected evolution trend according to the relative change amount; in the evolution trend, screening the continuous time domain that can cover the needs of the current round of action based on the allowable action amount, and then extracting the local trough interval where the load state changes from decreasing to increasing from the continuous time domain, and only when the remaining duration of the local trough interval from the candidate starting point is not less than the working time corresponding to the planned action amount of the current round, and the allowable action amount within the working time is not less than the planned action amount of the current round, the time domain from the starting point to the completion time of the planned action is determined as the candidate phase.

[0012] Preferably, the process of acquiring the access phase includes: converting the replenishment amount of the supplementary device and the usage amount of the configured access phase in each control cycle into pressure increment and pressure drop of the control cycle, respectively, based on the calibration relationship of the effective volume of the energy storage container, the medium state, and the corresponding pressure range; starting from the current pressure of the common execution source, superimposing the pressure increment and subtracting the pressure drop in each control cycle to form a pressure recovery trajectory covering each candidate phase; verifying whether the predicted pressure of the pressure recovery trajectory is always not lower than the working pressure boundary throughout the entire process of the candidate phase; and determining the candidate phase that meets the working pressure boundary and does not coincide with the configured access phase in the time domain as the access phase.

[0013] Preferably, the process of acquiring the opening and closing timing control command includes: embedding the access phase into the time axis of the reconstructed pressure recovery trajectory according to the order of the action period; determining the starting point of the access phase as the planned closing time, and taking the completion of the planned action quantity in this round or the measured body temperature reaching the safety boundary as the termination condition, and determining the time when the termination condition is met as the re-entry command time; forming the opening and closing timing control command by the planned closing time, the action time sequence of the working execution element within the phase, and the re-entry command time; after completing the current opening and closing timing control command, updating the allowable action quantity according to the evolution of the body temperature, and rearranging the opening and closing timing control commands that have not yet been executed according to the evolution of the common execution source pressure.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs a local reference within a time domain comparable to the production stage and the total drive pressure difference. It extracts the recovery response after re-injection from the transition in the total flow characterization, and correlates the recovery response with the action quantity according to the control cycle, establishing a relationship between the unit action quantity and the recovery response. Then, it determines the action requirements for the current cycle based on the expected time when the operational deviation reaches its upper limit, the deviation increase before the next operation, and the amount of incomplete action. Compared to fixed-experience control, this scheme can isolate the disturbances of natural evolution of operating conditions, allowing the action quantity to adaptively change with the unit's recovery capability and deviation degree, reducing insufficient recovery and excessive action, and improving the accuracy of demand determination and the stability of control.

[0015] 2. This invention establishes a safety boundary using the lower limit of temperature control and the protection margin of the opening and closing response hysteresis. It constructs a non-operational temperature baseline based on temperature evolution during the same period and selects a conservative trajectory by identifying the largest temperature margin consumption in historical operations to extrapolate the allowable operation amount before the safety boundary. Furthermore, it utilizes the propagation delay from upstream events to state transitions to migrate historical evolution trends and selects local load troughs from the time domain covering the current operation requirements. This scheme uses temperature safety margin, operational continuity, and load for phase screening, preventing operations from crossing the safety boundary or falling into high-load periods, thus improving the safety, feasibility, and production adaptability of candidate phases.

[0016] 3. This invention converts the replenishment amount and phase usage amount into pressure increments and pressure decreases, and reconstructs the pressure recovery trajectory of the common execution source cycle by cycle. The entry conditions are that the pressure within a phase is not lower than the working pressure boundary and that phases do not overlap. Entry phases are allocated according to the action period and pressure recovery order, forming control commands for closing, action, and re-engagement. After the commands are completed, the subsequent pressure recovery trajectory is reconstructed based on the measured pressure evolution, and unexecuted control commands are rearranged. This prevents pressure insufficiency and phase conflicts caused by multiple units using the same resources simultaneously, takes into account unit action periods, and ensures orderly connection and continuous feasibility of operations under shared execution resources. Attached Figure Description

[0017] Figure 1 A schematic flowchart of a method for controlling the opening and closing timing of a high-temperature blocking device based on production data is provided by the present invention. Figure 2 This is a schematic diagram of the logic flow for deducing the permissible amount of action provided by the present invention; Figure 3 A schematic diagram of the candidate phase screening logic flow provided by the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. Example

[0019] The high-temperature blocking device described in this embodiment refers to a blocking device installed between the controlled unit and the system aggregation space, having two working positions: conduction and blocking, and whose working medium temperature is higher than the temperature at which the medium undergoes phase change and condensation under the current humidity and pressure. The controlled unit refers to a unit arranged parallel to other similar units, sharing access to and from the aggregation space, and carrying the medium under the same driving pressure differential. Each controlled unit contains an execution element that takes the working medium from a common execution source and injects it back into the carrying element; the common execution source refers to an energy storage container and its replenishment device that supply the working medium to the execution elements of all controlled units. The opening and closing action of the high-temperature blocking device itself does not complete the operation; it merely causes the controlled unit to exit the carrying capacity, thereby creating a time domain available for the execution element to operate; this time domain is called the admission phase when determined during the layout phase, and is called the operation phase when actually occupied by the execution.

[0020] Due to the lag in mechanical execution, this paper distinguishes four moments: the shutdown command moment (the moment the controller issues the shutdown command); the shutdown completion moment (the moment the device is shut down and the feedback is set); the reactivation command moment (the moment the controller issues the reactivation command after the termination condition is met); and the reactivation completion moment (the moment the device is reactivated and the feedback is set). The admission phase begins with the planned shutdown completion moment; the operational phase begins with the actual shutdown completion moment and ends with the reactivation command moment; the pre- and post-response sampling for recovery is divided by the reactivation completion moment. Undefined shutdown and reactivation moments are not used throughout this paper.

[0021] The action quantities involved in this embodiment are divided into five items, and their order is as follows: Current action requirement, which refers to the requirement quantity retrieved from the unit response characterization based on the target recovery degree; Allowable action quantity, which refers to the upper limit of the action quantity allowed by the temperature safety boundary; Planned action quantity, which refers to the action quantity written into the opening and closing sequence control instruction after the current action requirement is limited by the allowable action quantity and verified by the supply conditions of the common execution source; Actual action quantity, which refers to the action quantity completed within the work phase; Incomplete action quantity, which refers to the difference between the current action requirement and the actual action quantity, including both the portion that was not included in the current plan after being limited by the allowable action quantity and the portion that was not executed due to the early termination of the work phase. The two are combined into the same state quantity for carry-over. The allowable action quantity is only used as an upper limit and should not be used as the target quantity for continued execution when the current action requirement is small.

[0022] The upstream production cycle time on which control is based refers to the temporal sequence of the upstream production process's operating status, composed of three types of process parameters: the operating and stopping status of upstream equipment, the switching signals of process stages, and the material feeding rate. All three types of parameters can be directly collected by field instruments or existing control systems. The upstream production cycle time does not contain any planned, scheduling, or instructional information. The time interval between the switching positions of two adjacent stages in the upstream production cycle time constitutes a production stage. Within the same production stage, the upstream material feeding rate and equipment operation combination remain unchanged. This property is the common premise for the validity of all subsequent extrapolations and transfers.

[0023] The measurement points configured accordingly include: the body temperature measurement point of each controlled unit, the medium temperature, humidity, and pressure measurement points on its feed side; the total flow rate measurement point and the total drive differential pressure measurement point on the system aggregation side; the pressure measurement point of the common actuator energy storage container, and the operating status and discharge parameters of the supplementary device; and the position feedback signal of each high-temperature blocking device drive unit. The three types of process parameters on the upstream production process side are read from the existing production control system. During the commissioning phase, a temporary pressure gauge is connected to the inlet side of the operating actuator, solely for obtaining the calibration value of the equivalent supply pressure drop. After calibration, it is removed and does not constitute a necessary measurement point for operation. All of the above measurement points are standard configurations; this method does not add any new actuators or change any equipment body.

[0024] Since the sampling period, filtering delay and timestamp of each measuring point are not consistent, the controller records the timestamp of each measuring point with a unified clock and uses the control cycle of the work execution element as the calculation grid: the continuous measurement quantity in the same grid is taken as the valid value that is closest to the grid time and does not exceed the allowable delay, and the state quantity is taken as the most recent valid state before the grid time; if any critical quantity has not been updated for more than one calculation grid, no new candidate phase is generated in this cycle, and the verified safe state is used.

[0025] At the two measuring points on the system convergence side, the controller does not directly use the total drive differential pressure. Instead, it takes the ratio of the system's total flow measurement to the system's total drive differential pressure measurement, referred to as the total flow characterization. The purpose of this ratio is to mitigate the common changes caused by the slow movement of the common induced draft fan's operating point, rather than eliminating this effect under all operating conditions. Accordingly, the controller only uses this characterization within the same production phase, during periods when the total drive differential pressure is within the historically comparable range for that phase and there are no other opening or closing actions. It does not assume, based on this, that the resistance relationship of each branch is linear under all operating conditions. The historically comparable range is determined by the measured distribution of the total drive differential pressure during periods of no action within that production phase, taking the upper and lower boundaries of its intermediate section.

[0026] Please see Figure 1 This invention provides a method for controlling the opening and closing timing of high-temperature blocking equipment based on production data. The technical solution is as follows: A unit response characterization is established based on the historical action quantities of each controlled unit and the recovery response after re-entry. The action duration and target recovery degree of each controlled unit are determined based on the evolution of operational deviations, and the current round of action requirements are retrieved from the unit response characterization. After aligning according to the propagation delay from upstream events to controlled units, the temporal evolution of temperature and load states is analyzed along the upstream production cycle. The allowable action quantity is determined based on the margin of the temperature state relative to the safety boundary. The allowable action quantity is used to limit the current round of action requirements to obtain the planned action quantity for this round. The time domain where the load state is at a local trough and its duration is not less than the operation time corresponding to the planned action quantity for this round is determined as the candidate phase. Based on the pressure recovery trajectory of the common execution source, verify whether the predicted pressure of the candidate phase throughout the entire process is always not lower than the working pressure boundary. The candidate phase that passes the verification and does not overlap with the configured access phase in the time domain is determined as the access phase, and the pressure recovery trajectory is reconstructed according to its usage relationship. According to the order of action deadlines, based on the successively reconstructed pressure recovery trajectory, non-overlapping access phases are configured for each controlled unit, forming and outputting opening and closing timing control commands. The opening and closing timing control commands are executed, with the closing time as the starting point of the working phase. When the measured body temperature reaches the safety boundary or the planned action amount of this round is completed, the first one to stop subsequent triggering and issue a re-entry command. The unit response characterization and the next round of action requirements are updated based on the actual action amount and the recovery response after re-entry.

[0027] Furthermore, the process of acquiring the unit response characterization includes: taking the time of re-entry as the boundary, within a continuous time domain in which the upstream production cycle has not been switched, there are no other opening and closing actions, and the total driving pressure difference is within the historical comparable range of this production stage, continuing the evolution of the total flow characterization before re-entry to form a local reference; determining the transition of the total flow characterization after re-entry relative to the local reference as the recovery response, and associating the recovery response with the actual action quantity of the corresponding operation phase according to the control cycle to form a unit response characterization that characterizes the correspondence between the action quantity and the flow recovery.

[0028] Specifically, taking the re-entry time of the controlled unit as the boundary, the controller searches backward for continuous periods within the same production stage where the total drive pressure difference is within a historically comparable range and there are no opening or closing actions of any controlled unit. Within these periods, the total flow characterization sequence is extracted, and its fluctuation amplitude is checked to see if it falls within the stable segment judgment bandwidth. This bandwidth is statistically obtained from the short-term fluctuation amplitude of the total flow characterization measured by the device during periods without any opening or closing actions, and its upper envelope value is taken. Its function is to eliminate readings distorted by natural fluctuations in incoming materials. The sequence that passes the test is extended after the re-entry time according to its change trend to form a local reference. After re-entry and the transition process has decayed, a stable segment is extracted from the total flow characterization sequence, and the difference between its reading and the value of the local reference at the same time is the recovery response of the phase of this operation. Under the aforementioned constraints, this reading is only relevant to the activated unit. The basis for this is that each controlled unit is arranged in parallel, sharing an inlet and outlet convergence space. Under the same driving pressure differential, the sum of their volumetric flow rates equals the total system flow rate. Therefore, the flow characteristics of each unit within this pressure differential range can be added together. Readings are not used if the conditions are not met. If the fluctuation amplitude of the intercepted segment exceeds the stable segment judgment bandwidth, or if the total driving pressure differential before and after reactivation exceeds the historical comparable range, this reading is invalidated and recorded, and not included in the unit response characterization.

[0029] The recovery response and the actual action quantity are stored in pairs in the unit response representation of the controlled unit. The action quantity referred to here is the cumulative number of operations performed by the work execution element within a work phase, recorded with the control cycle as the smallest division. The control cycle is determined by the sum of the duration of a single action of the work execution element and the minimum interval between two actions; both are taken from the technical documents of the execution element and verified with measured action records during the commissioning phase. The unit response representation uses the controlled unit as an index, storing the cumulative action quantity and its recovery response reading corresponding to each division. When the same cumulative action quantity obtains different readings in multiple rounds, the median value is taken. During retrieval, if the target value exceeds the coverage of the existing division, it is extended outward by one division according to the change trend between the two most recent divisions. This position is marked as the extrapolation position, and its corresponding action quantity is taken in the upward bias direction in this round to avoid underestimating the demand due to extrapolation.

[0030] The recovery response written into the unit response characterization is the increase in flow characterization brought about by this operation, not the absolute amount after the unit is re-engaged. The target recovery degree to be compared is itself the loss increment. Although the two have the same dimensions, their data roles are different. Comparing them in absolute terms would miscalculate the unit's existing carrying capacity as the result of this operation, thus underestimating the amount of action. Therefore, the controller obtains the stable segment reading after the unit is fully closed, using the closing position time of the same phase as the boundary. The increase in the reading on the re-engagement side relative to this reading is written as the recovery response.

[0031] The establishment of unit response characterization is divided into a table creation phase and an operation phase. After the device is put into operation, the table creation phase begins. For each controlled unit, a trial run is performed according to the action quantity scale from smallest to largest, constrained by safety and working pressure boundaries. At least two effective recovery responses of different action quantity scales are obtained before retrieval based on the target recovery degree is initiated. If there are fewer than two effective scales, no extension is made; the conservative action quantity confirmed in the equipment technical documents or debugging records is used, and data collection continues. During the operation phase, after each update of the unit response characterization, the controller checks whether the recovery response remains unchanged in the direction of increasing action quantity: if the later scale is lower than the previous scale and the difference exceeds the stability segment judgment bandwidth, the scale is marked as pending review, and this scale is not used in this round of retrieval; if the difference is a slight inversion and does not exceed the stability segment judgment bandwidth, it is treated as a value not higher than the adjacent effective scale. Extrapolation is only allowed to extend by one scale. If the target recovery degree cannot be covered after extension, the restricted handling described later is applied, and continuous extrapolation is not performed. If a controlled unit fails to obtain an effective recovery response for several consecutive rounds, the automatic arrangement of the unit will be suspended and a prompt to verify the measurement points and their arrival feedback will be output.

[0032] Furthermore, the process of obtaining the current round of action requirements includes: extending the evolution of the operating deviation formed by adjacent control cycles in segments along the upstream production cycle to obtain the estimated time when the operating deviation reaches the upper limit of the allowable operating range; reserving the shutdown response time and the time required for the current round of operation ahead of the estimated time, and determining the latest shutdown command time as the action deadline; taking the estimated value obtained by extending the evolution of the operating deviation to the completion time of the current round of operation as the deadline deviation state; determining the target recovery degree by the sum of the expected growth amount between the deadline deviation state and the earliest time when the controlled unit can obtain the next operation phase from the completion time of the current round of operation, relative to the upper limit of the allowable operating range; searching the unit response representation in order of increasing action amount to find the position where the recovery response first covers the target recovery degree; when there is an incomplete action amount, taking the position corresponding to the incomplete action amount as the search starting point, and determining the action amount corresponding to the search position as the current round of action requirements.

[0033] Specifically, the current operational requirements are derived from the system's current operational deviation. This operational deviation refers to the difference between the clean baseline flow characterization and the current total flow characterization, i.e., the cumulative decrease in the total flow characterization caused by all controlled units. The operational deviation is not characterized by the difference between the total driving pressure difference and the clean baseline pressure difference because the target recovery level derived from the operational deviation must be directly compared with the recovery response in the unit response characterization, and the recovery response is taken from the transition in the total flow characterization. If the operational deviation is taken as the pressure difference, the two have different dimensions, making it impossible to compare the recovery response to ensure it is not less than the target recovery level for the first time. The clean baseline flow characterization is measured during the period after the device is put into operation, when all controlled units have completed their work and are re-engaged. Thereafter, it is re-measured and updated each time all units have completed their work within a set number of cycles. The allowable operational range of the operational deviation is determined by the allowable decrease in the total flow characterization specified in the equipment's technical documents. The controller extends the operational deviation trend formed by adjacent control cycles within the current production stage. If a production stage switch occurs during the extension process, the extension continues according to the historical trend of the stage after the switch, starting from the switch time. The reason why this segmented extension is valid is that the upstream feeding rate and equipment operation combination remain unchanged within the same production stage, the load intensity entering the system is approximately constant, and the growth trend of the operational deviation is therefore approximately maintained within this stage; once the production stage is switched, the load intensity changes accordingly, and the trend must also be changed to the record of the stage after the switch.

[0034] The moment when the deviation from the above-mentioned direction extends to the upper limit of the allowable operating range is the moment when the unit can no longer wait. Since the shut-down action itself requires response time and the operation itself requires execution time, the action deadline is not taken at this moment, but is taken from the latest shut-down command time obtained after reserving the shut-down response time and the time required for this round of operation. The shut-down response time is taken as the historical measured upper envelope value of the equipment from the shut-down command time to the shut-down completion time. The time required for this round of operation is estimated by the product of the actual action quantity of the previous round and the control cycle, and is corrected after the planned action quantity of this round is determined. This correction is only performed once in a round of arrangement and is not iterated: the time obtained from the actual action quantity of the previous round is used as the initial value to obtain the action deadline, the target recovery degree and the action requirement of this round in sequence. After limiting the planned action quantity, it is used to replace the time required for the operation and the action deadline once each. If the remaining time of the candidate phase after correction is insufficient to accommodate the operation time, the requirement is not recalculated, but the planned action quantity is adjusted down according to the action quantity that the remaining time can accommodate, and the difference is treated as the uncompleted action quantity.

[0035] The expected value obtained by extending the operational deviation evolution to the completion time of this round of operations is the time-bound deviation state. The target recovery degree is not the difference between the time-bound deviation state and the upper limit of the allowable operating range—if this difference were taken, since the extension endpoint itself is defined as the upper limit position, the result would always be zero, and the retrieval would fall into zero action, which would contradict the purpose of recovering the operational deviation. The controller instead takes: the time-bound deviation state, plus the expected increase in operational deviation of the controlled unit from the completion time of this round of operations to the earliest time when it can obtain the next operational phase, and the amount by which the sum of the two exceeds the upper limit of the allowable operating range, which is the target recovery degree. The earliest time when the next operational phase can be obtained is determined by the later of the rotation interval determined by the lower limit of the number of operating units and the time required for the unit body temperature to recover to the point where it can operate again. With this definition, the goal of this round of recovery is to ensure that the operational deviation does not exceed the upper limit of the allowable operating range after recovery until the unit can operate again. When the sum of the two is not greater than the upper limit of the allowed operating range, the excess is not positive. In this round, no action requirements will be generated for this unit and no operation phase will be opened. In the next cycle, the deviation will be recalculated according to the updated operating deviation evolution.

[0036] The controller searches the unit response representation of the unit in ascending order of action quantity, and takes the division position where the recovery response is not less than the target recovery degree for the first time. The corresponding action quantity is the action requirement for this round. If there is an incomplete action quantity in the unit, that is, the difference between the planned action quantity and the actual action quantity of this round formed by the early termination of the previous operation phase due to reaching the safety boundary, the division position corresponding to the incomplete action quantity is used as the starting point for searching backward.

[0037] The operational requirements of this round determine the amount of action needed to recover from operational deviations, while the upper limit of the allowable amount of action within the temperature safety boundary is determined by the temperature margin of the controlled unit itself. The controller divides the candidate time period using the stage switching position of the upstream production cycle as the endpoint. For each candidate time period, it searches backwards for consecutive time periods that were in the same production stage before the start of that time period and had no opening or closing actions. It extracts the body temperature sequence from these periods and extends it into the candidate time period according to its trend of change, forming a no-action temperature baseline, which serves as a reference for measuring the temperature deviation caused by the operation. The corresponding safety boundary consists of two parts: the first is the lower limit of temperature control, which is retrieved from a pre-stored medium phase change temperature correspondence table. This correspondence table is pre-established under the premise that the applicable medium composition of the device is determined, covering the specified pressure range, and is expressed in terms of the output diameter of the humidity measuring point in this application. The pressure is uniformly based on the same standard, and both humidity and pressure are taken from the synchronous measurement values ​​on the feed side. When the measured value exceeds the range of the correspondence table or the medium composition changes, the automatic generation of allowable action quantity is suspended, and the conservative lower limit of temperature control confirmed by the equipment technical documents is used instead. The second is the protection margin, which is obtained by multiplying the response time required for the high temperature blocking device from receiving the re-entry instruction to the actual feedback setting by the rate of temperature drop of the body within the time. During the debugging stage, the response time is obtained by statistically analyzing the action records of no less than a set number of actions and taking its upper envelope value. The rate of drop is taken from the actual measured value of the last segment of the historical operation phase of the unit in the same production stage. The reason for setting a protection margin is that after the termination condition is met, the reactivation action still needs to be performed. During the time occupied by this action, the body temperature continues to drop. If the safety boundary is only the lower limit of temperature control, the measured temperature at the time of reactivation will definitely exceed the limit. After reserving this margin, the body temperature at the time of reactivation will still not be lower than the lower limit of temperature control.

[0038] The process of temperature advancing from the initial value of the candidate time period towards the safety boundary is depicted by the temperature margin consumption trajectory. This trajectory is taken from historical records: the controller selects historical operation phases in which the controlled unit is in the same production stage and the starting body temperature of the operation phase falls within the same body temperature scale interval. In each record, the temperature margin consumption amount at that moment is the temperature difference between the baseline temperature at no action and the measured body temperature. When the measured body temperature is not lower than the baseline temperature at no action, the consumption amount for that period is recorded as 0 and stored in pairs with the cumulative action amount executed at that moment. The action progress is represented by the cumulative action amount rather than its proportion of the planned action amount in this round, because the purpose of the temperature margin consumption trajectory is to extrapolate the allowable action amount, and the planned action amount in this round can only be obtained after the allowable action amount is determined. If the proportion is used as the independent variable of this trajectory, a circular dependency will be formed, and the extrapolation result cannot directly give the absolute number of actions. When there are multiple valid records under the same conditions, the one with the larger consumption amount is selected point by point, that is, the one that results in a smaller remaining temperature margin is selected point by point, so that the extrapolation is biased towards the safety side.

[0039] The temperature margin consumption is determined in the aforementioned direction because the body temperature during the operational phase is typically lower than the baseline temperature at which no action is performed. Subtracting the baseline from the measured value yields a negative value, with larger values ​​corresponding to smaller temperature drops, contrary to the intention of favoring a safer approach. The calculation is performed using the following single method: starting from the initial body temperature of the candidate period, the temperature margin consumption corresponding to each increment of the action quantity is taken. This consumption is then subtracted from the baseline temperature at the same time as the deviation from the baseline at which no action is performed. The result is the predicted body temperature at the completion of that increment. This consumption represents the total deviation accumulated up to that increment and is not added back periodically. The predicted temperature of each increment is compared to the safety boundary, and the calculation continues until the temperature first falls below the safety boundary. The number of actions that can be continuously performed before this point is the allowable number of actions for that candidate period. Regarding heat balance, once the controlled unit is placed in the blocking position, no high-temperature medium continues to pass through, and its heat input is interrupted. However, each time the operating element operates, it injects room-temperature operating medium taken from the common operating source into the unit. This medium absorbs heat and carries it out, resulting in a continuous heat outflow. The sum of these two factors makes the heat balance of the unit net outflow during the operating phase. However, due to the influence of residual high-temperature medium, shell temperature gradient, thermal inertia of measuring points, and environmental radiation, the measured temperature in individual control cycles may remain flat or rise briefly. Therefore, this method does not assume that each control cycle will monotonically decrease. The temperature margin consumption trajectory is taken from the actual measurement record, and the actual changes are retained for valid records that show flatness or short-term rise.

[0040] The aforementioned stable segment determination bandwidth, historical comparable range, number of action records required for response time statistics, number of records required for each scale interval, local trough determination quantile, and body temperature scale interval are all determined in the debugging records according to the following unified method for each production stage: Statistics are performed segment by segment using fixed-length inactive windows. First, faulty measurement point samples and cross-stage switching samples are removed. Then, the convergence condition is that the statistical results of several consecutive windows do not exceed the range corresponding to the measurement point resolution. The cumulative number of samples when the convergence condition is met is taken as the minimum number of records for that threshold. Each threshold is only updated when a new valid sample reaches a complete statistical window; it is frozen and not updated during anomaly handling. The physical meaning, data source, update triggering conditions, and failure conditions of each threshold are summarized and listed in the debugging records.

[0041] Furthermore, referring to Figure 2The process of obtaining the permissible action quantity includes: determining the safety boundary based on the temperature control lower limit corresponding to the medium state and the protection margin corresponding to the opening and closing response hysteresis; for each candidate time period divided by the stage switching position of the upstream production cycle, taking the body temperature evolution before the start of the candidate time period and in the same production stage as a local reference, and forming a non-action temperature baseline along the local reference; associating the temperature difference between the non-action temperature baseline and the measured body temperature during historical actions with the action process to form a temperature margin consumption trajectory; and sequentially deducing from the starting temperature of the candidate time period along the temperature margin consumption trajectory to determine the action quantity that can be continuously executed before reaching the safety boundary as the permissible action quantity.

[0042] Once the permissible action quantity is determined, the controller limits its action requirements for the current round: if the current round's action requirements are not greater than the permissible action quantity, the planned action quantity for the current round is the current round's action requirements; if the current round's action requirements are greater than the permissible action quantity, the planned action quantity for the current round is the permissible action quantity, and the difference is recorded as uncompleted action quantity and carried over to the next round. The permissible action quantity is never used as the execution target quantity for the current round under any circumstances. After the planned action quantity for the current round is determined accordingly, the selection of candidate phases is based on this quantity instead of the current round's action requirements: if the permissible action quantity is still required to cover the current round's action requirements, then when the current round's action requirements are greater than the permissible action quantity, no candidate phase can be formed. The current round cannot execute the limited portion of the work, and the carryover of uncompleted action quantity cannot proceed; the unit can only wait until the action deadline and be transferred to restricted processing. After changing the selection basis to the planned action quantity for the current round, the permissible action quantity will necessarily cover it, and the current round can execute the limited portion of the work. The difference is recorded as uncompleted action quantity and used as the starting point for the search in the next round.

[0043] Furthermore, referring to Figure 3 The process of acquiring the candidate phase includes: taking the upstream event formed by the stage switching signal or the change in feeding rate as the starting point, searching the historical records for the moment when the total flow and the body temperature show a corresponding turning point, and obtaining the propagation delay of the production stage; dividing the candidate time period along the stage switching position of the upstream production cycle, shifting the historical temperature state evolution and load state evolution within the same production stage according to the propagation delay to the corresponding candidate time period, and correcting the evolution trend according to the starting state of the candidate time period; in the corrected evolution trend, selecting the continuous time domain that can cover the needs of the current round of actions according to the allowable action amount, and then extracting the local trough interval where the load state changes from decreasing to increasing from the continuous time domain. Only when the remaining duration of the interval from the candidate starting point is not less than the working time corresponding to the planned action amount of the current round, the time domain from the starting point to the completion time of the planned action is determined as a candidate phase; when no candidate phase is formed, the current opening and closing state is maintained, and the phase is re-selected in the next production stage.

[0044] Specifically, changes in upstream process parameters do not arrive at the controlled unit at the same moment they occur; there is a propagation delay determined by the transport process. If the historical evolution is directly migrated at the same relative moment, the historical trough will be earlier or later than the current cycle, resulting in a systematic misalignment of the candidate phase. Therefore, the controller first obtains the propagation delay of the production stage: starting with the upstream event formed by the stage switching signal or change in the feeding rate, it retrieves the corresponding turning points of the system's total flow and body temperature from the historical record. The difference between the upstream event moment and the total flow turning point is the load propagation delay, and the difference to the body temperature turning point is the temperature propagation delay. The two evolutions are shifted according to their respective delays. The flow transport and instrument response processes differ from the body's thermal inertia; the two turning points do not necessarily occur simultaneously. The historical evolution is stored with the upstream event moment of its respective stage as the time origin, and the shift is applied only once. Multiple values ​​from the same production stage form its historical stable range. When the current value exceeds this stable range, the historical evolution is not used to generate a candidate phase; instead, re-acquisition is initiated, and the current open / closed state is maintained.

[0045] The selection of historical evolution data follows these rules: The production stage identifier is used as the primary index; the feeding rate at the start of the stage, equipment operating combinations, and the range of body temperature and total flow rate are used as matching conditions to select historical stages that match the current candidate time period. Temperature and load state evolution must be taken from the same historical stage and cannot be taken from different stages. When multiple valid records exist, representative change trends are formed based on relative times within the stage; records with distorted measurement points, intervention during the period, or premature end of the stage are all discarded. No candidate phase is generated if there are insufficient valid records.

[0046] The controller shifts the selected temperature and load state changes by propagation delay, then places them at the same relative time in the current candidate time period according to the relative time within each stage. The starting point of the placed trend is then replaced with the measured temperature and load state values ​​at the beginning of the current candidate time period. Subsequent time periods are derived sequentially based on the relative changes of the placed trend, thus preserving the shape of the trend while resetting its starting point, resulting in the corrected evolution trend. Here, temperature state refers to the sequence of measured values ​​from the temperature measurement points of the controlled unit, and load state refers to the sequence of measured values ​​from the total system flow measurement points. The permissible action quantity is calculated hourly along the corrected trajectory. A continuous time domain is defined as the uninterrupted time interval where the permissible action quantity remains no less than the current round's action requirement. Within this continuous time domain, the turning point where the load state changes from decreasing to increasing is retrieved. The intervals on either side of this turning point where the load state is below the local trough threshold are selected. The remaining duration of each candidate starting point within this interval is verified backwards. Only when the remaining duration is no less than the operating time corresponding to the product of the current round's planned action quantity and the control cycle, and the permissible action quantity within this duration remains no less than the current round's planned action quantity, is the time domain from that starting point to the completion time of the planned action considered a candidate phase. This threshold is determined statistically from the historical distribution of load states within the same production stage. Values ​​below a set percentile are selected from this distribution. The percentile is determined separately for each production stage during the plant's commissioning phase based on the fluctuation range of the load state. Production stages with larger fluctuation ranges use lower percentiles to avoid misjudging normal fluctuations as troughs. Without this remaining duration verification, the operation will extend into the load increase phase, and the control objective of the action falling within a trough cannot be achieved. If no candidate phase is formed within the current selection period, the controlled unit maintains its current open / closed state and is re-selected in the next production stage; if its action deadline is reached first during this period, it will proceed to the restricted handling described below.

[0047] Furthermore, the process of acquiring the admission phase includes: converting the replenishment amount of the supplementary device and the usage amount of the configured admission phase in each control cycle into pressure increment and pressure drop for that cycle, respectively, based on the calibration relationship of the effective volume of the energy storage container, the medium state, and the corresponding pressure range; starting from the current pressure of the common execution source, superimposing the pressure increment and subtracting the pressure drop for each control cycle to form a pressure recovery trajectory covering each candidate phase; verifying whether the predicted pressure of the pressure recovery trajectory is always not lower than the working pressure boundary throughout the candidate phase process; determining the candidate phase that meets the working pressure boundary and does not coincide with the configured admission phase in the time domain as the admission phase, and reconstructing the pressure recovery trajectory based on the usage relationship of the admission phase, so that the next controlled unit continues to make admission determination based on the reconstructed pressure recovery trajectory.

[0048] Specifically, the candidate phase only guarantees that the unit's own conditions are met; whether it can be configured as the admission phase depends on whether the common execution source can continuously supply power within that time domain. The discharge rate of the supplementary device and the amount taken by the operating execution element are both medium quantities and cannot be directly superimposed or subtracted from the pressure: pressure changes also depend on the effective volume of the energy storage container, the compressibility of the medium, the medium temperature, and the pressure level at the time of extraction. Therefore, the controller first converts the supplementary amount and the extraction amount of the configured operation for each control cycle into the pressure increment and pressure drop for that cycle, respectively, based on the calibration relationship of the effective volume of the energy storage container, the medium state, and the corresponding pressure range. Then, it updates the pressure for the next cycle starting from the pressure of the previous cycle, recursively forming a pressure recovery trajectory cycle by cycle. This calibration relationship is obtained segmented according to the pressure range during the commissioning phase; if the calibration relationship is not valid or the medium temperature exceeds the calibration range, the predicted trajectory is discontinued and recalculated using the measured pressure.

[0049] The consumption amount and working pressure boundary for each action are obtained as follows: During the commissioning phase, assuming the common execution source pressure is within the same pressure range and the supplementary device is operating in the same state, the evolution of the source end pressure is recorded when there is no consumption and when a single execution element is rated to consume. Combining the minimum inlet working pressure and single consumption amount recorded in the execution element's technical documentation, and using a temporarily connected inlet-side pressure gauge, the equivalent supply pressure drop of the corresponding supply pipeline for the controlled unit is obtained. The working pressure boundary, which is the sum of the minimum inlet working pressure and the equivalent supply pressure drop, is stored separately for each controlled unit.

[0050] The order in which controlled units enter the judgment process is arranged according to their action deadlines, with those having smaller action requirements in the same round being prioritized. For the currently judged unit, the controller first takes a copy of the pressure recovery trajectory. Based on the planned action amount in this round, the pressure drop calculated by the candidate unit itself is added to the copy periodically according to the action time sequence within the candidate phase. Then, it is checked whether the predicted pressure after addition remains above the working pressure boundary throughout the entire phase and whether it coincides with the configured admission phase. Adding before checking is necessary because once the candidate unit is executed, the state being checked changes; otherwise, a situation might occur where the check passes but the pressure drops below the boundary during execution. Units that pass both checks are designated as admission phases, and this copy serves as the reconstructed trajectory for the next unit to judge. If all candidate phases of a unit fail the check, the unit is not configured in this round, its action deadline remains unchanged, and it is placed first among units with the same deadline in the next judgment cycle. Since the admission determination is performed unit by unit in the above order, and each determination has eliminated the overlap with the configured admission phase, after the configuration is completed, each admission phase will not overlap with each other in the time domain, and there is no need to perform overlap resolution.

[0051] Furthermore, the process of acquiring the opening and closing timing control command includes: embedding each access phase into the time axis of the reconstructed pressure recovery trajectory according to the order of action deadlines; determining the starting point of the access phase as the planned closing time, taking the completion of the planned action quantity in this round or the actual body temperature reaching the safety boundary as the termination condition, determining the time when the termination condition is met as the re-injection command time, and forming the opening and closing timing control command by the planned closing time, the action time sequence of the working execution element within the phase, and the re-injection command time; after completing the current opening and closing timing control command, updating the allowable action quantity according to the evolution of the body temperature, and rearranging the opening and closing timing control commands that have not yet been executed according to the evolution of the common execution source pressure.

[0052] Specifically, each access phase is embedded into the timeline of the reconstructed pressure recovery trajectory according to the order of its action duration. The starting point of the access phase is the planned shutdown time of the controlled unit; the termination condition is the completion of the planned action in this round or the measured body temperature reaching the safety boundary, whichever occurs first. The moment the termination condition is met is the moment the command is reissued. The termination quantity here is the planned action quantity in this round, not the allowable action quantity: the allowable action quantity is the maximum executable quantity under the temperature safety boundary. If it is still used as the termination quantity when the action requirement in this round is small, the number of actions will exceed the control requirement, and the correspondence between action quantity and recovery response in the unit response characterization will deviate from the actual requirement. The shutdown command time is obtained by reserving the shutdown response time before the planned shutdown time, and the controller issues the shutdown command at this time. The shutdown response time is taken from the historical measured upper envelope value. This reservation ensures that the actual shutdown time is no later than the planned shutdown time, but does not necessarily make them equal. Therefore, deviations are handled as follows: if the actual shutdown is earlier than planned, the unit remains in the blocking position and waits. The first operation trigger is no earlier than the planned shutdown time, ensuring that the access still falls within the verified pressure time domain. If the actual shutdown lags behind by more than one control cycle, the remaining low-valley time, allowable action amount, and pressure trajectory are re-verified starting from the actual shutdown time, and the planned action amount is lowered to the verified amount. If the remaining time domain is insufficient for one action, the current operation is canceled and a reactivation command is issued. The shutdown command time, the planned shutdown time, the action time sequence of the operation actuator within the phase, and the reactivation command time together constitute the opening and closing timing control command of the controlled unit.

[0053] The controller sequentially issues opening and closing timing control commands to each controlled unit according to the time axis. After the closing command is issued, the start point of the operation phase is set by the position feedback of the drive unit. Within the phase, the operation execution element is triggered sequentially according to the control cycle. After each trigger, the measured value of the body temperature and the measured value of the common execution source pressure are read. When the body temperature reaches the safety boundary, subsequent triggering is immediately stopped and a restart command is issued. Similarly, when the planned action quantity for this round is completed but the body temperature has not yet reached the safety boundary, a restart command is also issued. In both cases, the uncompleted action quantity is recorded as the difference between the required action quantity and the actual action quantity for this round. When the requirement has been fully executed, it is recorded as zero. Limiting deficit and execution deficit are carried forward together and no second item is set up for accounting. After being reinstated, the controller obtains the recovery response as described above and writes it into the unit response characterization in pairs with the actual action quantities. It then recalculates the unit's operational deviation evolution, action duration, and target recovery degree to obtain the action requirements for the next round. Simultaneously, it takes the temperature difference between the baseline temperature without action within the current operation phase and the measured value of the body temperature, updates the temperature margin consumption trajectory of the unit in the corresponding production stage and the corresponding starting temperature interval according to the cumulative action quantity, and updates the allowable action quantity for its subsequent candidate time periods accordingly. Furthermore, it corrects the calibration relationship between pressure increment and pressure drop based on the measured evolution of the common execution source pressure within the current phase, and rearranges the unexecuted opening and closing timing control commands accordingly. These four updates flow back to the unit response characterization, action requirements, allowable action quantity, and pressure recovery trajectory, respectively, so that the arrangement of the next round is based on the measured results of this round.

[0054] If the action deadline of a controlled unit has arrived but a phase that has passed the judgment has not yet been formed, and its candidate phase sequence is not empty, the controller will forcibly configure the phase with the largest allowable action quantity in its candidate phase sequence as the admission phase if all of the following conditions are met: the unit's body temperature measurement point, common execution source pressure measurement point, system total flow measurement point, and position feedback signal are all valid; the supplementary device is in operation; the planned action quantity obtained after adjusting the allowable action quantity of this phase, after the pressure drop converted by its use is included in the trajectory copy, the predicted pressure of this phase throughout the entire process is not lower than the working pressure boundary; the actual planned execution quantity is used for verification because the verification conclusion cannot cover the actual use when the measured quantity is less than the actual execution quantity. After forced configuration, the planned action quantity of this round is adjusted down according to the allowable action quantity of this phase, and the part that is not completed in this round is recorded as the incomplete action quantity, which will be used as the starting point for the next round of action requirement retrieval. If the candidate phase sequence is empty or the verification fails, forced configuration will not be performed, and the following restricted handling will be applied.

[0055] If any of the above conditions is met (e.g., a critical measuring point is invalid, the supplementary device is shut down, the feedback on closing is not set after timeout), or the predicted pressure is lower than the working pressure boundary after verification of the planned actions, the controller cancels the closing command. If the unit has already been closed, subsequent operation triggering is stopped and a reactivation command is immediately issued. The unexecuted portion of the planned actions in this round is included in the incomplete actions and carried over to the next round. The executed portion is still counted in the actual actions and is not recorded as incomplete again. At the same time, the manual review status is output. Meeting the safety boundary takes precedence over meeting the action deadline, and the action deadline cannot be used as a basis for closing beyond the safety conditions. For restricted actions, the uncompleted portion is recorded in the incomplete actions and carried over, and is used as the starting point for priority replenishment in subsequent rounds. If the temperature margin or supply is continuously insufficient, this portion may span multiple rounds. The controller outputs its carry-over status and the reason for the restriction in each round.

[0056] In this embodiment, the action time falls within a local trough of the load state, and this range is sufficient to accommodate the operation time corresponding to the planned action volume of this round. During the period when the controlled unit exits the load, the load increment borne by the remaining operating units is at a lower level in the same production stage, which can be verified by the measured ratio of the total system flow to the number of operating units. The operation phase terminates when the body temperature reaches the safety boundary, and the safety boundary includes a protection margin covering the restart response time. The measured body temperature at the restart time is not lower than the lower temperature control limit retrieved from the medium phase change temperature correspondence table based on the current humidity and pressure of the medium. The installation position of the body temperature measuring point is determined during the commissioning phase by comparing multiple points in the load-bearing element area with the readings at the same time, ensuring that the readings are not higher than those of the control points. Therefore, when the temperature control lower limit is met, no phase change condensation occurs on the surface of the bearing element, which can be verified by the synchronous recording of the body temperature measuring point and the medium state measuring point; the entire process of the access phase is checked against the working pressure boundary through the pressure recovery trajectory, and each phase does not overlap in the time domain. The inlet pressure of the operating element is not lower than its minimum working pressure throughout the entire operation phase, and the operation is carried out under rated conditions, which can be verified by the recording of the common execution source pressure measuring point; the action requirements of this round are determined by the unit response characterization retrieval. The action quantity of each controlled unit is differentiated according to its own measured action quantity and the correspondence between flow recovery and no longer takes the same set value. The flow recovery quantity corresponding to the unit action quantity increases and the number of actions per unit time decreases, which can be verified by the historical records and action logs of the unit response characterization. Example

[0057] This embodiment provides a supplementary implementation method based on Embodiment 1. In Embodiment 1, the current round of action requirements and action deadlines are derived from system operational deviations, which are the total amount caused by all controlled units. Based on this, the target recovery level is determined for each controlled unit, which is equivalent to assuming that each unit shares the burden equally. At the same time, Embodiment 1 only depicts the process of temperature margin being consumed within the working phase, without constraining its regeneration after re-entry. Controlled units may be re-entered into the candidate phase before their body temperature has recovered, and their allowed action quantity will be far lower than the current round of action requirements. Once the working phase is opened, the cost of exiting the load has already occurred, and the current round of operation cannot be completed. This embodiment addresses the above two points, and the remaining steps are the same as in Embodiment 1.

[0058] Furthermore, the process of acquiring the unit response characterization also includes: taking the closing time as the boundary, within a continuous time domain where the upstream production cycle has not switched, there are no other opening and closing actions, and the total driving pressure difference is within the historical comparable range of this production stage, continuing the evolution of the total flow characterization before closing to form a local reference on the closing side; when the change in the total flow of the system before and after closing is within the comparison bandwidth, the drop in the total flow characterization after closing relative to the local reference on the closing side is determined as the pre-action contribution of the controlled unit; when the change exceeds the comparison bandwidth, the current reading is invalidated and the previous pre-action contribution of the controlled unit is used; the recovery response of the same operation phase is determined as the post-action contribution of the controlled unit, and the increase in the post-action contribution relative to the pre-action contribution is determined as the actual effectiveness of the operation phase, and associated with the actual action amount according to the control cycle; according to the proportion of the pre-action contribution of each controlled unit relative to its summation value, the system target recovery degree derived from the operation deviation is allocated to the target recovery degree of each controlled unit.

[0059] Specifically, the shut-off action and the restart action represent two symmetrical reading opportunities in the total current characterization, and the latter was used in Example 1. The controller uses the shut-off feedback setting time as a boundary to search forward for continuous periods within the same production stage where the total drive pressure difference is within the historically comparable range of that stage and there are no controlled unit opening or closing actions. Within these periods, a total current characterization sequence is extracted, and its fluctuation amplitude is checked to see if it falls within the stable segment judgment bandwidth. This bandwidth is obtained in the same way as in Example 1, by statistically analyzing the short-term fluctuation amplitude of the total current characterization measured by the device during periods without any opening or closing actions, and taking its upper envelope value. The sequence that passes the test is extended after the shut-off time according to its trend, forming a local reference on the shut-off side. After the transition process has decayed completely after shut-off, a stable segment is extracted from the total current characterization sequence, and the drop in its reading relative to the local reference on the shut-off side represents the contribution amount of the controlled unit before its action. The drop is only related to the closed unit, and its basis is the same as the transition in Example 1: each controlled unit is set up in parallel and shares the inlet and outlet collection space. Under the same driving pressure difference, the sum of their volumetric flow rates is the total flow rate of the system. Therefore, the flow characteristics of each unit can be added within the pressure difference range.

[0060] The closing-side reading requires an additional condition because the closing action redistributes the total system flow, and the magnitude of this redistribution is greater than the local disturbance during reactivation. When the resistance of the load-bearing element is predominantly viscous, the flow rate and pressure difference are approximately linear, and the flow characteristics of the unit are approximately unaffected by the flow velocity. However, if the change in total flow rate before and after closing is too large, the inertial component of the resistance cannot be ignored, the readings on both sides are not under the same comparison conditions, and the drop value is no longer equal to the flow characteristics of that unit. Therefore, the controller has a comparison bandwidth, the value of which is determined by the measured relationship between the flow rate and pressure difference of the load-bearing element near the rated flow rate. The deviation of the flow characteristics from its rated value is taken to be no more than the flow rate change corresponding to the measurement resolution of the measuring point. This relationship is obtained during the commissioning phase using synchronous flow rate and pressure difference records under different flow conditions and stored separately for each controlled unit. The reading is valid when the change in the total system flow rate before and after closing is within this bandwidth; if it exceeds this bandwidth, the current reading is invalidated, and the contribution amount before the previous action of that unit is used. If a controlled unit exceeds the comparison bandwidth for a set number of consecutive rounds, the controller will schedule the unit's action to the candidate phase with the smallest load state fluctuation amplitude to obtain a valid reading. This will prevent the unit's pre-action contribution from remaining at the old value for a long time, which would distort subsequent judgments. The set number of rounds is determined by the number of rounds in which the unit's pre-action contribution can still support an admission decision without effective updates. The lower quartile of the number of rounds between two consecutive valid readings in the historical record is taken.

[0061] The recovery response of the same operational phase is the contribution of the controlled unit after its action. Both the contribution before and after the action are absolute quantities of the current flow characterization of the unit at the corresponding moment, and they cannot be added together. The increase of the contribution after the action relative to the contribution before the action is the actual effectiveness of this operational phase. The controller stores the actual effectiveness and the actual action quantity of this operation in pairs in the unit response characterization of the unit according to the control cycle. Its storage, mid-level processing, non-monotonic verification, and extrapolation rules are the same as in Example 1.

[0062] The sharing relationship of operational deviations among controlled units is established as follows: The controller aggregates the pre-action contributions of all controlled units at the same effective time. It first checks whether the difference between this aggregated value and the system's total current characteristic during the same period is within the stability judgment bandwidth to confirm that the readings of each unit were obtained under the same comparison conditions and without omissions. After passing the check, the target recovery degree derived from the operational deviation is allocated to each unit according to the proportion of each unit's pre-action contribution relative to the aggregated value. This proportional allocation, instead of directly correlating the pre-action contribution with the operational deviation, is used because the former represents the unit's current absolute current contribution, while the latter represents the decrease in the system's relative cleanliness baseline; both have the same dimension but different data roles. The proportion obtained through proportional allocation is a dimensionless coefficient, ensuring consistent dimensions before and after allocation. Units with a larger share receive a larger target recovery degree, while units with a smaller share are no longer assigned operating phases incompatible with their status. If any unit uses the old value in this round, or if the aggregated check fails, the sharing relationship is not updated in this round; the sharing relationship that passed the check in the previous round continues to be used, and a re-check is performed in the next round.

[0063] Further, the process of obtaining the action period includes: taking the decrease in the contribution amount after the action of the previous operation phase relative to the contribution amount before the action of the next operation phase for the same controlled unit between two adjacent operation phases, dividing it by the interval between the two readings to obtain the current decay rate of the controlled unit in that production phase; determining the decay rate of that phase only using samples where the start and end readings are both in the same production phase and no phase switching occurs in between, and not writing cross-phase samples into the decay rate record of any production phase; extending the contribution amount before the action of the controlled unit in segments according to the decay rate of the current production phase to the expected time when it drops to the lower limit of the unit's allowable current flow, reserving the shutdown response time and the time required for this round of operation ahead of the expected time, and determining the latest shutdown command time as the action period of the controlled unit; determining the order of entry into the admission judgment according to the action periods of each controlled unit.

[0064] Specifically, the readings between two adjacent readings carry the unit's own attenuation information. The contribution after the operation of the previous phase is taken from the moment the unit is re-engaged after completing its operation, representing the high value of the current flow characterization within that cycle. The contribution before the operation of the next phase is taken from the moment it is closed again, where the current flow characterization has decreased due to accumulation, representing a low value. The difference between the two is the decrease, which, when divided by the interval between the two readings, is the current flow attenuation rate. The direction must be taken in the above order; if the difference is taken in reverse, the result with the opposite sign will be obtained, and the extrapolation direction of the operation period will be reversed accordingly.

[0065] The decay rate must be taken separately for each production stage, and only samples from the same stage can be used. During the interval, the unit is under load, and its decay rate depends on the load intensity during that period. The load intensity is approximately constant within the same production stage but changes across stages. Two adjacent readings only give the total decay for the entire interval. When the interval spans multiple production stages, the duration of each stage alone cannot distinguish how much decay each stage caused. Allocating the decay by duration is equivalent to assuming the decay rate is the same for each stage, which contradicts the purpose of taking values ​​in stages. Therefore, the controller only uses samples whose start and end readings are within the same production stage and where no stage switch occurs in between; cross-stage samples are not recorded in any stage's record. If cross-stage samples are necessary, a separate operating phase should be arranged at the stage switch location to obtain segmentable intermediate contribution readings, based on which the decay rate for each stage is calculated. This arrangement itself is still constrained by safety and operating pressure boundaries. If the number of effective samples for a production stage is insufficient, the decay rate of the adjacent production stage with the closest load intensity is used, and this stage is marked as awaiting supplementary sampling.

[0066] The lower limit of the allowable flow rate of a unit is determined by the contribution amount of the controlled unit after its operation in a clean state and the allowable decrease in flow capacity specified in the equipment technical documents. The contribution amount after its operation in a clean state is obtained when the unit completes its first operation and is put back into place after the device is put into operation; thereafter, the contribution amount is remeasured and updated when the actual effective amount of a certain operation phase of the unit does not exceed the measurement resolution of the measuring point and the actual action amount of that phase is not less than that of the previous phase; the difference between the effective amounts of two adjacent phases is close to zero, which only indicates that the recovery amount tends to be stable and both can be significantly positive at the same time, and is not used as a criterion for the unit to have reached the upper limit of recovery. The controller extends the unit's pre-operation contribution to the expected time when it drops to the unit's allowable current limit, based on the decay rate of the current production stage. It then reserves a shutdown response time and the required time for this round of operation from that expected time forward. The latest shutdown command time obtained is the unit's operation period. The two reserved times are from the same source as in Example 1. The shutdown response time is the historical measured upper envelope value of the equipment from the shutdown command time to the shutdown completion time. The required time for this round of operation is estimated based on the product of the actual operation amount of the previous round and the control cycle, and corrected after the planned operation amount for this round is determined. Each controlled unit determines its entry order for access judgment according to its operation period; if the operation periods are the same, the unit with the smaller operation requirement for this round is prioritized.

[0067] Furthermore, it also includes the determination of heat reserve regeneration: starting from the moment of re-entry into position, the evolution of the body temperature back to the bearing state temperature is collected. The first moment when the change in body temperature for multiple consecutive control cycles does not exceed the measurement resolution of the body temperature measuring point and the body temperature does not change downward again during the continuous period is determined as the heat reserve regeneration completion moment of the controlled unit. Before the heat reserve regeneration is completed, the controlled unit does not participate in the selection of candidate phases. When the action period of the controlled unit reaches the heat reserve regeneration completion moment before the time of its action period, the allowable action amount is re-estimated based on the measured body temperature at that time and participates in the selection. The planned action amount for this round is determined according to the allowable action amount, and the part that is not completed is recorded as the incomplete action amount. Based on the correspondence between the historical regeneration completion time and the medium temperature during the regeneration period, it is used to predict the moment when each controlled unit participates in the selection in the subsequent candidate period.

[0068] Specifically, after the controlled unit is repositioned, the high-temperature medium resumes flow, and the heat input recovers accordingly. The unit temperature rises from the low value at the end of the operating phase to the bearing state temperature, with the rate of recovery decreasing over time. The controller collects the unit temperature cycle by cycle from the moment of repositioning, forming a record of the recovery evolution. The heat reserve regeneration mentioned here and the temperature reserve consumption described in Example 1 refer to the same reserve in two directions: during the operating phase, this reserve is carried away by the injected room-temperature operating medium, and after repositioning, it is re-accumulated under the condition that the high-temperature medium resumes flow.

[0069] The criterion for completing regeneration does not consider the change in temperature within a single adjacent control cycle; the recovery process gradually approaches the operating temperature, and its rate will not return to zero, so it must be defined by a criterion; while the change in a single adjacent cycle is very small, possibly due to measurement quantization, transient plateaus, or random noise, and does not indicate that the temperature has stabilized and recovered. Judging based on this would cause the unit to participate in the screening prematurely, overestimating its thermal margin. Therefore, the controller requires that the change in the body temperature over multiple consecutive control cycles does not exceed the measurement resolution of the body temperature measuring point, and that the body temperature does not change downward again during this continuous period. The first moment that meets the above conditions is taken as the time when thermal margin regeneration is completed. The number of consecutive cycles only includes records of units that naturally enter the stable phase during previous regenerations and do not subsequently issue a shutdown command prematurely due to arrangement. The shortest duration of the stable phase is divided by the control cycle and rounded down; records truncated by arrangement reflect past arrangement results rather than physical recovery patterns and are not used. The obtained number of cycles shall not be less than the number of control cycles corresponding to the filtering delay of the body temperature measurement point. If it is lower than this value or there are insufficient records, the number of cycles shall be used and marked as to be supplemented. Zero or one shall not be used.

[0070] Before the regeneration of the heat margin is completed, the controlled unit does not participate in the selection of candidate phases. If the unit's action deadline arrives before its regeneration completion time, the controller does not allow it to wait indefinitely. Instead, it re-deduces the temperature margin consumption trajectory based on the measured body temperature at that time to obtain the allowable action quantity at that moment and participates in the selection. The planned action quantity for this round is determined according to the allowable action quantity. The part not covered in the action requirements of this round is recorded as the incomplete action quantity, which is used as the starting point for the next round of action requirement retrieval. This handling follows the same priority as the restricted handling in Embodiment 1, that is, the satisfaction of the safety boundary takes precedence over the satisfaction of the action deadline.

[0071] To enable the completion time of heat recovery regeneration to be predicted in advance rather than determined retrospectively, the controller uses the medium temperature division as an index, storing the median value of the regeneration completion time for each division. The division width is determined according to the measurement resolution of the medium temperature measuring point, and the number of records required for each division is determined using the same convergence condition as the thresholds in Example 1. During prediction, the regeneration completion time of the corresponding division is retrieved based on the predicted medium temperature value of the candidate time period to determine the earliest time for each controlled unit to participate in the screening within that time period. If the predicted value falls outside the existing division, it is extended by one division according to the trend of the most recent division, and the extended position is marked as the extrapolation position. Its duration is taken in the upward bias direction in this round to avoid underestimating the regeneration time required.

[0072] The pre-action and post-action contributions are generated by the inevitable shutdown and restart actions, without increasing detection methods. Except in cases where additional work phases are arranged at stage switching positions to obtain segmentable intermediate contributions, the number of actions is not increased; however, the additional work phases arranged in such cases are counted in the unit's action count. The target recovery degree is allocated according to the proportion of flow characterization for each controlled unit, and the action duration is determined separately according to the flow decay rate of each unit. The order of admission judgment is arranged according to the predicted action duration of each unit, thus reducing the number of times restricted treatment is entered and the amount of incomplete actions. Heat margin regeneration judgment ensures that work phases are only activated when the temperature margin is sufficient to support the planned action amount for this round. Actions whose durations precede regeneration completion are performed on the limited portion, increasing the completion rate of each work phase and reducing the total number of work phases required under the same operational deviation control level. All of the above can be verified by total flow characterization records, unit temperature records, and action logs.

[0073] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for controlling the opening and closing timing of a high-temperature blocking device based on production data, characterized in that, include: Based on the historical action volume of each controlled unit and the recovery response after reactivation, a unit response characterization is established, and the action requirements for this round are determined in combination with operational deviations. The safety boundary of the upstream production cycle is analyzed to meet the temperature control state and the opening and closing response state. Combined with temperature extrapolation, the amount of action that can be performed before reaching the safety boundary is determined as the allowable action amount. The time domain that covers the current action demand and the load state is at a local trough is determined as the candidate phase. Based on the pressure recovery trajectory of the public execution source, the supply conditions of the candidate phase for the current action demand are verified, and the candidate phase that passes the supply verification is determined as the admission phase; According to the action period and pressure recovery sequence, each controlled unit is configured with a non-overlapping access phase to form and output opening and closing timing control commands. Execute the opening and closing timing control command, terminate the current operation phase when the measured temperature reaches the safety boundary or the current round of action is completed, and update the unit response characterization and the next round of action requirements based on the actual amount of action and the recovery response after re-entry.

2. The method for controlling the opening and closing timing of a high-temperature blocking device based on production data according to claim 1, characterized in that: The process of acquiring the unit response characterization includes: taking the time of re-entry as the boundary, within a continuous time domain in which the upstream production cycle has not been switched, there are no opening or closing actions, and the total driving pressure difference is within the historical comparable range of the production stage, continuing the evolution of the total flow characterization before re-entry to form a local reference; determining the transition of the total flow characterization after re-entry relative to the local reference as the recovery response; and associating the recovery response with the actual action quantity of the corresponding operation phase according to the control cycle to form a unit response characterization that characterizes the correspondence between the action quantity and the flow recovery.

3. The method for controlling the opening and closing timing of a high-temperature blocking device based on production data according to claim 2, characterized in that: The process of obtaining the current round of action requirements includes: extending the evolution of the operating deviation formed by adjacent control cycles in segments along the upstream production cycle to obtain the estimated time when the operating deviation reaches the upper limit of the allowable operating range; reserving the shutdown response time and the time required for this round of operation ahead of the estimated time, and determining the latest shutdown command time as the action deadline; taking the estimated value obtained by extending the evolution of the operating deviation to the completion time of this round of operation as the deadline deviation state; determining the target recovery degree by the sum of the expected growth between the deadline deviation state and the earliest time when the controlled unit can obtain the next operation phase from the completion time of this round of operation, relative to the upper limit of the allowable operating range; searching the unit response representation in order of increasing action quantity to the position where the recovery response is first not less than the target recovery degree; when there is an incomplete action quantity, taking the position corresponding to the incomplete action quantity as the search starting point, and determining the action quantity corresponding to the search position as the action requirement for this round.

4. The method for controlling the opening and closing timing of a high-temperature blocking device based on production data according to claim 3, characterized in that: The process of obtaining the permissible action quantity includes: determining the safety boundary based on the lower limit of temperature control corresponding to the medium state and the protection margin corresponding to the opening and closing response lag; for each candidate time period divided by the stage switching position of the upstream production cycle, using the body temperature evolution before the start of the candidate time period and in the same production stage as a local reference, and forming a no-action temperature baseline along the local reference; in each control cycle during the historical action period, using the temperature difference between the no-action temperature baseline and the measured body temperature as the temperature margin consumption, and associating the temperature margin consumption with the action process, when there are multiple valid records under the same conditions, taking the one with the larger consumption point one by one to form a temperature margin consumption trajectory; and sequentially deducing from the starting temperature of the candidate time period along the temperature margin consumption trajectory to determine the action quantity that can be executed before reaching the safety boundary as the permissible action quantity.

5. The method for controlling the opening and closing timing of a high-temperature blocking device based on production data according to claim 4, characterized in that: The process of obtaining the candidate phase includes: taking the upstream event formed by the stage switching signal or the change in the feeding rate as the starting point, searching the historical records for the moment when the total flow and the body temperature show a corresponding turning point, and obtaining the propagation delay of the production stage; shifting the historical temperature state evolution and load state evolution within the same production stage and with the same stage start condition as the current candidate time period according to the propagation delay to the corresponding candidate time period, and replacing the starting point of the migrated trend with the starting state of the candidate time period, and sequentially deriving the corrected evolution trend according to the relative change amount; in the evolution trend, screening the continuous time domain that can cover the needs of the current round of action based on the allowable action amount, and then extracting the local trough interval where the load state changes from decreasing to increasing from the continuous time domain, and only when the remaining duration of the local trough interval from the candidate starting point is not less than the working time corresponding to the planned action amount of the current round, and the allowable action amount within the working time is not less than the planned action amount of the current round, the time domain from the starting point to the completion time of the planned action is determined as the candidate phase.

6. The method for controlling the opening and closing timing of a high-temperature blocking device based on production data according to claim 1, characterized in that: The process of acquiring the access phase includes: converting the replenishment amount of the supplementary device and the usage amount of the configured access phase in each control cycle into pressure increment and pressure drop of the control cycle, respectively, based on the calibration relationship of the effective volume of the energy storage container, the medium state, and the corresponding pressure range; starting from the current pressure of the common execution source, adding pressure increment and subtracting pressure drop for each control cycle to form a pressure recovery trajectory covering each candidate phase; verifying whether the predicted pressure of the pressure recovery trajectory is always not lower than the working pressure boundary throughout the entire process of the candidate phase; and determining the candidate phase that meets the working pressure boundary and does not coincide with the configured access phase in the time domain as the access phase.

7. The method for controlling the opening and closing timing of a high-temperature blocking device based on production data according to claim 6, characterized in that: The process of acquiring the opening and closing timing control command includes: embedding the access phase into the time axis of the reconstructed pressure recovery trajectory according to the order of action deadlines; determining the starting point of the access phase as the planned closing time, and using the completion of the planned action quantity in this round or the measured body temperature reaching the safety boundary as the termination condition, and determining the time when the termination condition is met as the re-injection command time; forming the opening and closing timing control command by the planned closing time, the action time sequence of the working execution element within the phase, and the re-injection command time; after completing the current opening and closing timing control command, updating the allowable action quantity according to the evolution of the body temperature, and rearranging the opening and closing timing control commands that have not yet been executed according to the evolution of the common execution source pressure.

Images